http://www.hardtecs4u.com/reviews/2002/fsb_166/
here what it says
Babel Fish Translation, In English:
Intro No question, for which Athlon XP with 166 MHz FSB stands before the door and will in Baelde are introduced. Naturally the question, which will bring this at increased output by the increased range of the processor with itself, is quite interesting, but not the decisive point, why we with this topic to have been concerned. The Overclocker fan municipality concerns itself practically already quite a while with the topic 133 MHz to 166 MHz FrontSideBus and many operates AMD processors within the range of this FSB. for one while circulates rumors in the Web that with some plates a "nose" would be present, why hardly considerable increases in output are to be registered with the rise of the FSB to 166 MHz. Only recently also my colleague novel Fischer in the report of the EPoX Ep-8kä2+ had itself to surprise vs. Chaintech 7VJL Apogee at it. These points, and feedbacks of various users caused us to regard us some plates in this connection once more near.
We would like to however already point out in the apron that the later following results in the long run no guarantor for the achievement of the coming new AMD processors can be, because in the long run are present us no such CCU. We made only a over clocking procedure of an at present usual AMD CCU with specified FSB of 133 MHz. How BIOS versions of Motherboards will possibly develop themselves further during introduction of a AMD Athlon XP 2700+ still and whether it applies to consider still appropriate defaults on the part of AMD or the chip sentence manufacturers, remains being waiting definitely.
Ranges Already one while appears with present AMD processors of the FrontSideBus as new bottle neck in the system. The past models with a specified FSB of 133 MHz (266 MHz GDR) make a maximum processor range possible of 2.1 GB/s after the following computation: 133 MHz * 64 bits * 2 packets per clock = 17024 Mill. Bit/s = 2128 Mill. Byte/s Practically according to the identical formula however also the computation of the memory range behaves, so that one likewise receives memory module here with a DDR266 (PC2100) a maximum range from 2,1 GB/s. Thus one would have the most optimal basis between processor and memory range in the theory. And evenly also in the theory an at present current AMD Athlon XP processor might not profit to memory with a range from a DDR333 (PC2700) from 2,7 GB/s. This however only in the theory, because in practice showed up that the VIA KT333 chip set with support for the DDR333 storing and pool of broadcasting corporations the AMD processors again to approx.. 5 % increased output to draw could do. Why this presents itself in such a way in practice, we described in our Speicherroundup
"most optimal conditions with memory accesses presupposed, one reaches module 2100 MB/s memory range in a theoretical view for example with a PC2100 GDR sdram. Optimally in this connection means however all operational sequence in smooth interaction, described before, (PAGE hit, Burst length, bank Interleaving, etc.). In the best case one reaches its in practice however only 95%, thus approx.. 2000 MB/s memory throughput, which to reaching end of page the field of a memory cell or the PAGE Size limitation to due is. As in a book, it is not possible also in sides of a memory cell to read infinitely without Umzublaettern. This sheet procedure always costs a certain time, until one can read on. Evenly those 5% maximally are it, the AMD processor based platforms á la VIA KT333 with 133 MHz GDR FSB by the employment of DDR333 (PC2700) memory modules with 166 MHz GDR storing act straight still for itself to work out can - one considers that the processor evenly possesses also only a maximum range of 2.1 GB/s. Here thus the FrontSideBus represents a new bottle neck."
If we proceed now from a rise of the FrontSideBus on 166 MHz, then shown calculation formula the range of the CPUs on maximally 2.7 GB/s increases upward, and already identical environments may be expected once fundamentally presupposed in the theory that in connection with DDR333 memory a performance increase should result here. Thus an increase of the FSB is a quite welcoming point on the part of AMD.
all I did was use the translator and coppyed and pasted
http://babelfish.altavista.com/tr
Facebook
Twitter
Instagram